Do we need an investment function for China?

 

Posted by on

Krugman has just published a post that addresses the economic situation in China. He thinks (as many other economists) that we should not be very worried about the possible side effects of China on the rest of the world. Well, the Chinese economic performance is already hitting many economies (mainly through commodity prices) around the world (has anyone said Brazil?), so maybe the “price effects” that Krugman mentions are larger than what he thinks they are.

But here I don’t want to attack Krugman (because overall he seems quite mild about Chinese prospects), but rather I would like to comment very briefly on a report mentioned by him and published by Willem Buiter at City. The report is worth reading, because it makes the case for a global slowdown driven by the (under)performance of the Chinese economy.

There are many sentences in the report I would largely agree with. For instance, when talking about why China’s government debt burden is not as rosy as it seems, Buiter mentions that:

The soaring non-financial private sector debt burden and the matching soaring banking and shadow banking sector balance sheets suggest that a future financial rescue of systemically important and/or politically well-connected insolvent private entities and SOEs by the central government is likely. This could seriously strain even the fiscal capacity of the central government.

However, few pages later, one can find the following remark:

Very little equity funding of capital expenditure takes place in China, and with future profitable investments likely to be found in sectors and industries very different from those in the past, average Tobin’s q (the ratio of market capitalization of existing capital to the current reproduction cost of capital) is bound to be less that marginal Tobin’s q (the ratio of the NPV of future profits on new capital expenditure to the current reproduction costs of capital). And it is marginal Tobin’s q that drives investment.

Not quite. And not because the theoretical argument is wrong in its own terms (which it is, because in the neoclassical model that Buiter is having in mind, all that matters is marginal q: higher marginal q values will imply higher levels of investment, being the difference between marginal and average q given by adjustment costs and other “imperfections”, so mentioning average q for the investment argument is simply noise), but because in advanced capitalist economies with proper financial markets (or at least the most advanced ones on earth), many empirical studies have found the poor performance of “q models” in predicting investment performance (see here, for instance). Many reasons can be adduced for such a poor performance: marginal q is not observable (so managers have a hard time using it for investment decisions), stock markets are not perfectly efficient and stock prices can carry some noise and not only fundamental information (so managers are misled by market prices), and so on. Anyway, these considerations are for developed economies (economies where one would expect q theory to pass with flying colours), and even in these economies the q theory is not very promising. But in China, who cares? Investment has been clearly driven by government policy, so even more reasonable investment theories (cash-flows, capacity utilization, etc.) will not apply either in the Chinese case (personally, I would bet that q has not even mentioned once in the Chinese government as rationale for investment decisions).

This does not mean that Tobin’s q, as an economic indicator, does not have any use at all. It does, but not as a driver of investment (as we explain here), but rather as a financial indicator of the relative dearness of the stock market, a use that has already been advocated by many financial market practitioners (and that I elaborate and substantially expand here for the Chinese case).

So, keep simple the investment drivers in the Chinese case.

Leave a Reply

Your email address will not be published. Required fields are marked *